End Stage Renal Disease (ESRD) treatment has come under increasing scrutiny in terms of the outcomes achieved, cost, and quality of care. To compare ESRD treatments and their outcomes, instruments that accurately capture and account for baseline or initial patient differences, or casemix, are required. This application proposes a training and research program designed 1) to support the training and development of a physician ESRD outcomes researcher and 2) to advance the state of the art in dialysis casemix adjustment methods. The candidate has extensive background in clinical nephrology as well as exposure through recent research to casemix adjustment of dialysis outcomes. The proposed training program will provide intensive exposure to state of the art methods in survival modeling, outcomes measurement, and casemix adjustment. The goal of the proposed research projeCt is to develop comprehensive and clinically meaningful casemix models for dialysis patients that control for baseline characteristics, and to use these models as tools to study dialysis outcomes. Research efforts will be applied to two of the United States Renal Data System special studies databases, the Case Mix Severity Study and the Casemix Adequacy Study. To achieve the proposed goal, the candidate will explore how the factors that determine casemix for dialysis patients may interact to influence outcomes, by extending these models to hospitalization as an outcome, by validating these factors among incident and prevalent dialysis patient populations. He will demonstrate the use of casemix models through application of developed models to determine whether racial differences in dialysis outcome remain after careful casemix adjustment. Finally, he will extend modeling of casemix to consider quality of life as a predictor of outcomes. Methods developed under this proposal will be generalizable and applicable to the study of other ESRD outcomes, such as quality of life, and to evaluating other questions of interest to health care policy makers, such as differences in ESRD outcomes according to sex.